Many commercial and government entities receive input data in paper format. Credit card applications, license applications, and tax returns are examples. Although input techniques such as automated telephone systems and Web-based utilities are increasingly provided as an input alternative, the volume of input made in paper format is still substantial. For instance, a bank or other financial institution may receive more than 10,000 credit card applications in paper format in a single day. Manual entry of this volume of data into data processing systems may be cost prohibitive.
A common technique for processing such a large volume of input forms involves automated or semi-automated conversion of the completed paper forms to electronic format. In a typical approach, paper forms are first scanned by a digital image scanner to yield an electronic bitmap image. The image is then converted to text via Optical Character Recognition (OCR) software (for reading machine printed characters) and Intelligent Character Recognition (ICR) software (for reading hand written characters). Not all data is correctly interpreted, however, due to functional limitations in scanning apparatus and recognition software. It is therefore common practice to employ data entry operators for the purpose of correcting errors or omissions resulting from the automated conversion process. For an example of such a system and method, see U.S. Pat. No. 5,054,096 issued to Beizer on Oct. 1, 1991.
Known approaches have several drawbacks and limitations, however. One issue not addressed by existing technology is how to manage the cost and cycle time associated with data entry staff. Another limitation is a failure to recognize that it may be advantages to process some applications differently than others, according to a variety of factors.
In sum, existing systems and techniques for converting information on paper forms to electronic data have not adequately managed the conversion process. The resulting lack of efficiency, and other drawbacks, limit the utility of such systems for entities receiving a high volume of input data in paper format.